home · mobile · calendar · colloquia · 1999-2000 · 

Colloquium - Singh

On-line Path Planning for Mobile Robots Operating in Vast Outdoor Environments
Robotics Institute, Carnegie Mellon University

Mobile robots operating in vast outdoor unstructured environments such as found on the surface of planets, start with coarse grained and incomplete maps at best. The robots must deal with new objects found during traversal finding new routes to the goal as necessary while maintaining reasonable traveling speeds. Path planning in such environments must thus be incremental to accommodate new information and must use efficient representations. In this talk I will discuss recent results in path planning using an efficient data structure called a "framed quadtree" and an optimal algorithm to incorporate knowledge of the environment as it is incrementally discovered. In particular, I will show the difference in performance when the robot starts with no information versus when it starts with partial information about the world. As expected, starting with partial information is better than starting with no information. However, in many cases, partial information results in performance that is almost as good as starting out with complete information, while the computational cost incurred is significantly lower.

Our system has been tested in simulation as well on an autonomous rover equipped with local obstacle avoidance based on stereo vision. We are in the process of porting these results onto a rover at NASA designed to explore Martian environments. Results from both simulation and real experimentation with our rover, Bullwinkle, will be presented.

Dr. Sanjiv Singh is a Scientist at the Robotics Institute, Carnegie Mellon University. His research interests include mobile robot navigation, machine learning and novel imaging devices. He is cited as co-inventor on ten US patents related to position guided navigation for mobile robots. Dr. Singh received a BS in Computer Science from the University of Denver, a MS in Electrical Engineering from Lehigh University, and a PhD in Robotics from Carnegie Mellon in 1995. He was a NSF Fellow at the Mechanical Engineering Laboratory in Tsukuba, Japan in 1992 and a Visiting Researcher at the University of Sydney in 1999.

Dr. Singh is available to speak with students interested in the Masters and Doctoral programs in Robotics at Carnegie Mellon.
Refreshments will be served afterwards in ECCS 122.
Hosted by Michael Mozer.

Department of Computer Science
University of Colorado Boulder
Boulder, CO 80309-0430 USA
May 5, 2012 (14:13)